Algorand’s solution to the blockchain trilemma

The Blockchain Trilemma

Most blockchains suffer from a trilemma when decentralization, scalability and security cannot be ensured on the blockchain at the same time.

Firstly, decentralization is defined as that property of the blockchain where all stake holders have access to the same amount of resources.
Secondly, scalability where the blockchain is able to process an ever increasing number of transactions within the smallest acceptable “deterministic” time.
Thirdly, any security transaction wherein the blockchain network is able to process the transaction without the possibility of any form of attack.
Most blockchains in today’s world suffers from one if not two of these issues. For example, bitcoin’s blockchain mining is highly concentrated. It is well known that scaling is also a significant problem with bitcoin’s blockchain – since the maximum number of transactions per second processable by the network has been limited by the block-size, and mining algorithm combined.

Algorand – a Blockchain that supports transactions of all types, including smart contracts, promises to be a solution for all three, through a unique and clever “true proof of stake” algorithm. Silvio Micali, the ACM Turing award winner and MIT Computer Science professor and his students are behind this phenomenal idea that accomplishes all three together.

Algorand accomplishes a proof of stake through a three stage process – which  doesn’t depend only on how much stake a miner has on the network. The protocol that Algorand uses is called the Byzantine Agreement protocol. BA not only satisfies some additional properties, but is very fast. Roughly said, its binary-input consists of 3 steps in which one of the participants sends a message ot all other players. The network is complete and synchronous,  where only those who are online at any given instant of time can participate.

In the paper the describes Algorand, Chen and Micali describe a unique mechanism for describing how blocks are generated.  More on this in a future post.